SV40 is a small virus but in its approximately 5,200 base pairs is enough information either to productively grow in a permissive cell, or to convert a normal cell into a transformed one. The conversion is usually gentically stable enough to generate clonable lines of transformed cells. Some of these lines lose so much of their normal growth control that they acquire the capacity to grow as tumors. Since the total DNA sequence of SV40 is known, it is somewhat surprising that the mechanism of action of SV40 in maintaining the various transformed states in vitro is not at all well understood. Also, and probably not coincidentally, precisely which SV40-encoded molecules are necessary for maintenance of any of these states is an unsettled issue as well. The first question cannot be easily asked without a study of the second. We propose here to ask both, through an interacting set of projects directed at the sequence-protein-function relationships of the SV40 transforming genes. We will initially test the hypothesis that variant non-lytic forms of large T are essential for full expression of the transformed phenotype. Our specific approaches are: (1): Directed deletion of SV40 DNA and analysis of lytic function. (2): Mapping and recovering SV40 from within transformed cells. (3): Transfection/transformation with novel SV40 DNA sequences. (4): Properties and activities of variant T antigen in lytic and transformed cells.